← Back to list
Project & Org Management
#이해관계자#권력관심격자#현저성모델#PMBOK#변화관리
Last updated · 2026-11-17

Stakeholder Management

1. Overview

A. Definition

A set of continuous management activities that identify and analyze every individual, group, or organization (stakeholder) that affects or is affected by a project or initiative, and manage their expectations and influence to secure active engagement. It is not mere "complaint handling" but a core Performance Domain of project success, on par with scope, schedule, and quality.

The essence of stakeholder management is bringing people and politics into the scope of project planning. Many IT projects are judged failures despite being technically complete — not because of defects in the code, but because of "people problems" such as non-cooperation from business units, withdrawal of executive support, or user resistance to change. No matter how faithfully requirements are implemented, if the key users who raised those requirements ultimately shun the system, the project delivers no value for the investment. Stakeholder management begins from the recognition that non-technical factors determine project success or failure, systematically revealing who influences the project and deliberately securing their support by design.

A key principle here is that stakeholder management is not a one-time event but a cyclical activity spanning the entire life cycle. Rather than cataloging stakeholders once at initiation and being done, new stakeholders emerge as the project progresses, existing stakeholders shift their attitudes from support to resistance (or vice versa), and the power landscape itself changes. For example, if a sponsoring executive is replaced through a reorganization, the top priority target changes overnight. The core mindset of stakeholder management is therefore to repeatedly run the identify→analyze→engage→monitor cycle and recalibrate dynamically.

B. Background and Necessity

Traditional project management focused on the technical control of the "triple constraint" of scope, schedule, and cost. However, as large SI, public informatization, and digital transformation projects proliferated, it became clear that a complex web of interests — involving multiple departments, partners, regulators, and citizens — determines the real success or failure of a project. In the public sector especially, external stakeholders who sit outside the project yet wield powerful influence, such as oversight agencies, civic groups, and the legislature, can reverse a project's direction. In the face of such complexity, the passive stance of "just gather the requirements well" is no longer valid.

Accordingly, PMI's PMBOK established "Project Stakeholder Management" as a separate knowledge area in its 5th edition in 2013, and in the 7th edition in 2021 it elevated "Stakeholder" to one of the 12 principles and the first of the 8 Performance Domains. This formalized the premise that effective collaboration with stakeholders is a precondition for all project activity. In short, stakeholder management has shifted in status from "nice to have" to "fail without it," and from a professional-engineer perspective it has become a subject to be treated with the same weight as architecture and governance design.

2. The Stakeholder Management Process and Overall Structure

Stakeholder management consists of four activities that are both sequential and cyclical. Each activity is not performed linearly just once; rather, monitoring results trigger renewed identification and analysis, forming a feedback loop. The structure diagram below shows this cycle along with the outputs and relationships of each stage.

flowchart TB
  subgraph CYCLE["Stakeholder management life cycle"]
    I["Identify<br/>build stakeholder register"]
    A["Analyze<br/>assess power-interest, salience"]
    P["Plan<br/>engagement strategy, comms plan"]
    M["Manage/Monitor<br/>sustain relations, track attitudes"]
    I --> A --> P --> M
    M -->|"detect new stakeholders, attitude shifts"| I
  end
  CHARTER["project charter,<br/>organizational context"] --> I
  M --> OUT["outcome: active support,<br/>reduced resistance, goal agreement"]
  style I fill:#e8f0fe,stroke:#2f6fed,stroke-width:2px
  style OUT fill:#e8f5e9,stroke:#34a853,stroke-width:2px

In the Identify Stakeholders stage, every entity with a stake in the project is found without omission. The output is the Stakeholder Register, recording each stakeholder's name, organization, role, expectations, influence, and interests. The most common failure at this stage is identifying only the "visible key parties" and missing hidden stakeholders (e.g., operations, the security review team, external auditors). Since poor identification starts all subsequent analysis with a gap, the principle is to sweep as broadly as possible using brainstorming, expert judgment, and templates from similar projects.

The stakeholder register is not a mere name list but the primary material for analysis. It typically has columns for identifying information (name, organization, title), assessment information (expectations, requirements, influence, attitude), and classification information (internal/external, friendly/adversarial, power-interest quadrant), and because it contains sensitive political assessments, access rights are controlled. Since the quality of the register determines the quality of the analysis, one must pursue both "without omission (identify)" and "accurately (assess)" at once.

In the Analyze stage, the identified stakeholders are evaluated by criteria such as power, interest, influence, and attitude to set priorities. Since the same resources cannot be invested in every stakeholder, the purpose of analysis is to decide where to concentrate. For example, in a large public project where 50 stakeholders are identified, treating them all identically is impossible, so analysis must differentiate the allocation — such as "5 manage closely, 8 keep satisfied, 20 keep informed, 17 monitor." The power-interest grid and salience model described below are the representative techniques of this stage.

In the Plan Stakeholder Engagement stage, an engagement strategy and communication plan are drawn up to move each stakeholder from its "current state" to its "required state." For example, to move a resisting key user to a "supportive" state, one designs what message to deliver, through which channel, and how often. In the Manage/Monitor stage, the plan is executed while relationships are sustained, and shifts in stakeholder attitudes and the emergence of new stakeholders are continuously tracked and fed back into identification and analysis.

3. Stakeholder Identification and Analysis Techniques

A. The Power/Interest Grid

The most widely used classification technique is the Power/Interest Grid proposed by Mendelow in 1981. Stakeholders are divided along two axes — "influence over the project (Power)" and "interest in the project (Interest)" — and placed into four quadrants. The grid is powerful because, in a reality of limited resources, it presents at a glance a differentiated strategy for "who deserves how much effort." Treating someone high in both power and interest the same as someone high only in interest is a waste of resources and a shortcut to relationship failure.

Specifically, the four quadrants demand different management strategies. The High power, High interest (Manage Closely) group holds the project's fate; sponsoring executives who approve the budget and business owners belong here. They require the most time, close consultation, and inclusion in decision-making. The High power, Low interest (Keep Satisfied) group normally shows little interest but can halt a project once dissatisfied — security, audit, and regulatory departments are typical. Because they stay uninvolved normally and intervene forcefully when problems arise, one manages them to a "no complaints" state by reporting only key decisions in a timely manner.

By contrast, the Low power, High interest (Keep Informed) group, like the business practitioners who will actually use the system, has limited influence but deep interest in the project. They can become the epicenter of change resistance or, conversely, a powerful base of support, so one wins them over with ample information and communication. The Low power, Low interest (Monitor) group is watched with minimal effort, while keeping an eye on whether a change of circumstances moves them into another quadrant.

flowchart TB
  subgraph GRID["Power/Interest Grid"]
    direction TB
    Q1["High power, Low interest<br/>Keep Satisfied<br/>e.g., audit, regulators"]
    Q2["High power, High interest<br/>Manage Closely<br/>e.g., sponsor, business owner"]
    Q3["Low power, Low interest<br/>Monitor<br/>e.g., indirectly related units"]
    Q4["Low power, High interest<br/>Keep Informed<br/>e.g., business users"]
  end
  Q2 -->|"top-priority resources"| TOP["project success"]
  Q1 -->|"block risk"| TOP
  style Q2 fill:#fdecea,stroke:#d93025,stroke-width:2px
  style Q4 fill:#e8f0fe,stroke:#2f6fed,stroke-width:1px

B. The Salience Model

What complements the power-interest grid's limitation — its two-axis simplicity can miss subtle configurations — is the Salience model of Mitchell, Agle & Wood (1997). This model classifies stakeholders by the combination of three attributes: Power, Legitimacy, and Urgency. Power is the ability to impose one's will; legitimacy is whether the claim is socially and contractually appropriate; urgency is how time-sensitively the claim demands immediate attention. The "salience" (degree of attention) a manager should grant varies with how many of the three attributes are held.

The combinations of the three attributes yield seven types. The Latent groups holding only one attribute (power only = Dormant, legitimacy only = Discretionary, urgency only = Demanding) have low salience. The Expectant groups holding two attributes (power+legitimacy = Dominant, legitimacy+urgency = Dependent, power+urgency = Dangerous) have medium salience. In particular, the Dangerous group, holding power and urgency but no legitimacy, warrants special caution because it can mobilize coercive means such as strikes, shows of force, or anti-system rallies. The Definitive group holding all three attributes has the highest salience and must be addressed immediately and with top priority. For example, in the wake of a data breach, a data protection authority becomes a definitive stakeholder possessing power (sanctioning authority), legitimacy (legal mandate), and urgency (demand for immediate remediation) all at once.

flowchart TB
  POW["Power"]
  LEG["Legitimacy"]
  URG["Urgency"]
  POW --> DORM["Dormant<br/>power only"]
  LEG --> DISC["Discretionary<br/>legitimacy only"]
  URG --> DEM["Demanding<br/>urgency only"]
  POW --> DOMI["Dominant<br/>power+legitimacy"]
  LEG --> DEP["Dependent<br/>legitimacy+urgency"]
  URG --> DANG["Dangerous<br/>power+urgency"]
  DOMI --> DEF["Definitive<br/>power+legitimacy+urgency<br/>top priority"]
  DEP --> DEF
  DANG --> DEF
  style DEF fill:#fdecea,stroke:#d93025,stroke-width:2px
  style DANG fill:#fef3e8,stroke:#ed8f2f,stroke-width:1px

C. The Stakeholder Engagement Assessment Matrix

The final analysis tool is the Stakeholder Engagement Assessment Matrix (SEAM), which makes visible the gap between each stakeholder's current and desired engagement levels. Engagement levels are usually divided into five stages: "Unaware → Resistant → Neutral → Supportive → Leading." The matrix marks each stakeholder's current position (C) and desired position (D) on the same row, revealing that the larger the distance between the two, the more engagement effort is needed. For example, if a key business team lead currently in a "Resistant" state must be raised to "Supportive," the specific activities to close that two-cell gap are reflected in the engagement plan. The value of this technique lies in converting abstract "relationship improvement" into a measurable goal, allowing the effectiveness of engagement strategy to be tracked.

Stakeholder Unaware Resistant Neutral Supportive Leading
Sponsoring executive C D
Business team lead C D
Operations unit C→D
Security review team C D

In the example above, the sponsoring executive need only be raised one step from supportive to leading, whereas the business team lead must be moved two steps from resistant to supportive, requiring more intensive persuasion and engagement activity. In this way the three techniques (power-interest grid, salience model, engagement matrix) are complementary: the grid captures the big picture of resource allocation, the salience model refines subtle power configurations, and the engagement matrix sets concrete execution goals.

4. Stakeholder Engagement Strategies and Practical Cases

If identification and analysis answer "where to concentrate," then the engagement strategy answers "how to move them toward support." The key is tailored communication matched to each stakeholder's type and engagement gap. For executives who are manage-closely targets, concise face-to-face reports covering KPIs, risks, and decision items are effective, while for business users who are keep-informed targets, change-management-oriented briefings, training, and early involvement (user workshops) work well. The same facts must vary in depth and format by audience.

A representative practical case is the problem of business resistance in a next-generation system transition project. Business users accustomed for years to familiar legacy screens tend to perceive a new system as an "uncomfortable imposition" and resist. Here, a "big-bang" approach of simply throwing a completed system at them almost always fails. Successful cases commonly used strategies that ① involve key users as Change Champions from the design stage to instill ownership, ② let them experience small wins through pilots and beta operation, and ③ build trust through hands-on floor walking and rapid feedback incorporation. In short, turning stakeholders from "management targets" into co-designers made the difference between success and failure.

Conversely, the typical failure case is neglecting the high-power, low-interest group (e.g., information security, legal) until they apply the brakes at the last moment. It is not uncommon for a security review near the end of development to flag a serious personal-data processing defect, forcing a full redesign. This is the result of regarding such a department as a "keep satisfied" target that should have received early review at each key decision point, yet deferring communication under the pressure of immediate development progress. This case shows that missing an influential "quiet stakeholder" is the most expensive mistake.

Stakeholder type Recommended engagement strategy Key outputs/channels
Manage closely (high power, high interest) Joint decision-making, regular steering committee Executive reports, steering committee
Keep satisfied (high power, low interest) Timely reporting of key decisions, early risk sharing Milestone reports, review consultations
Keep informed (low power, high interest) Change management, training, early involvement Briefings, user workshops
Monitor (low power, low interest) Minimal communication, status monitoring Announcements, newsletters

5. Deep Dive: The Agile/Digital-Transformation Context and the PMBOK 7 Perspective

In traditional (predictive) projects, stakeholder management was "plan-centric" — drawing up a register at initiation and communicating per plan. However, in an agile (adaptive) environment, the nature of stakeholder engagement changes fundamentally. Agile keeps the customer and Product Owner embedded in the team, or draws stakeholders continuously and directly into the development process through sprint reviews and demos. That is, stakeholder management is not a separate process but is built into the development rhythm itself, building trust by showing working deliverables every 2–4 weeks and taking immediate feedback. This lowers resistance far more effectively than "report and get approval through documents."

The PMBOK 7th edition, published in 2021, formalized this change in particular. The 7th edition was wholly restructured from a process/knowledge-area center to a principle and Performance Domain center, raising "Stakeholder" to one of the 12 principles ("engage stakeholders effectively") and the first of the 8 Performance Domains. The point is that it no longer mandates specific techniques (register, grid) but tells you to focus on the "outcome" of a productive relationship with stakeholders. The success of this performance domain is measured by "a productive relationship throughout the project" and "stakeholder agreement on objectives." This is the perspective — look at whether trust and collaboration actually formed rather than whether tools were used — that is also emphasized in professional-engineer examinations.

In practice, one example where this perspective worked is a large public platform construction project that involved local-government and related-agency staff in biweekly sprint reviews from the outset to jointly review working screens. Where the prior approach of repeated document approvals saw business complaints pour in only just before launch, those complaints now surfaced in small amounts each sprint and were resolved early, greatly reducing last-minute rework and resistance. This shows how effective a shift in engagement — turning "reported-to stakeholders" into "reviewing stakeholders" — is for building trust.

In the context of digital transformation and data governance, the scope of stakeholders expands further. In AI and personal-data projects, one must consider external stakeholders such as data subjects (citizens) and the data protection authority, and even the indefinite many affected by an algorithm. Recently, reflecting this, there is a clear trend of expanding stakeholders from a shareholder center to a broad stakeholder center from an ESG and social-responsibility perspective, which affects both corporate governance and project governance.

6. Considerations and Implications

  1. Stakeholder management must be combined with risk management. Resistance from influential stakeholders is itself a top-tier project risk. Therefore, stakeholder analysis results should be linked to the risk register, reflecting people-centered risks such as "withdrawal of a key executive's support" and "collective business resistance" into identification and response plans. Managing only technical risks while omitting stakeholder risks leaves the most fatal failure factor outside of control.

  2. Do not miss the "quiet but powerful" stakeholder. The most expensive mistake is neglecting the normally non-intervening, out-of-mind high-power, low-interest group (audit, security, regulators) until they apply the brakes at the last moment. These should be preemptively managed to a "no complaints" state through early review at key decision points, which prevents the enormous cost of rework and schedule delay.

  3. Engagement must shift from "management" to "co-creation" to be effective. Leaving stakeholders as passive report recipients breeds resistance, but involving them as parties in design and validation (nurturing champions, user workshops, agile demos) creates ownership and turns resistance into support. Combining this with change management (ADKAR, Kotter's 8 steps) to design an "awareness→desire→knowledge→ability→reinforcement" flow is the way to reduce trade-offs.

  4. Institutionalize dynamic reassessment. The stakeholder landscape changes frequently through reorganizations, personnel moves, and shifts in public opinion. Leaving the register from initiation as a "fixed document" quickly diverges from reality, so a procedure to update the register and engagement matrix at each regular steering committee and phase end should be built into governance.

  5. Presuppose ethics and transparency. Stakeholder management must be "building trust through honest communication," not "manipulation." Gaining short-term support by distorting or concealing information from a particular group collapses trust in the long run and invites greater resistance. From a professional-engineer perspective, one must recognize that transparent disclosure and fair arbitration of conflicts of interest are the foundation of sustainable stakeholder management.

  6. Operate in integration with other knowledge areas and governance. Stakeholder management is not an isolated activity but works interlocked with communications management (channels, cadence), scope management (reflecting requirements), and change management (consulting affected parties). In particular, clearly designing each stakeholder's decision authority and escalation path within the project governance structure (steering committee, operating committee) allows conflicts, when they arise, to be arbitrated by a pre-agreed procedure rather than by improvisation.

References


In one line: Stakeholder management is a life-cycle-wide cyclical activity that identifies and analyzes every party affecting or affected by a project (power-interest grid, salience model, engagement matrix) to turn resistance into support — so central to non-technical success factors that PMBOK 7 elevated it to the first Performance Domain and principle — and in agile/digital environments it evolves beyond "management" into "co-creation."